Niall Douglas wrote:
Myself and Boost have had a confrontational relationship in the past, but in my opinion (you and other will disagree), y'all after arguing heavily with me at the time then a few years later went ahead and quietly implemented almost everything I suggested. So I'm good with Boost at the present time - I spoke, you listened, you eventually implemented much of it. Rock on!
I have to admit that I can't at the moment think of any feature about which I argued with you when you proposed it, and then went ahead and quietly implemented later. It's true that you had already proposed https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2018/p1196r0.html but I'm not sure I actually knew that when I implemented this, and even if I did, I definitely don't remember arguing with you about it. https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2018/p1197r0.html is - maybe - something you had proposed earlier, but not to Boost and not in this form; the only discussion you were involved in about that (that I know about) is the SG14 one, in which I hadn't participated. I only read the paper https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2017/p0824r0.pdf but I think that at this point I've already heard enough complaints about the std::string use in <system_error> that I'd probably have written 1197 even without reading the SG14 discussion summary. It's also true that system::result is very similar to outcome::result (and std::expected), but I don't think I argued with you about having that, either. I'm perfectly willing to concede that you were there first. It's the arguing against part that I have no recollection of.